Whoops; I messed up when I added numbers. It alternates adding numbers and punctuation until all the numbers are unlocked, but mixed things up so it kept trying (and failing) to unlock an 11th number instead of an 11th punctuation symbol. Should be fixed now. (Though as Doug noted, you can also click on symbols to unlock them manually.)
There's a bunch of stuff like that that I just haven't gotten around to (see my previous comments about this not being intended for general use yet ). Though since you requested it and fonts are typically pretty straightforward, I might get to it sooner.
This is on my (mental) wishlist, but it's not clear to me that it's worth the effort. My intuition is that it's more useful to have the bias towards real-world bigram frequences than to give you ones you have trouble with (e.g., qu and ae are much more common than qz, so should show up more often). Plus, in order to know how slow you are on a bigram, you need to type it; you're not likely to see qz unless there's a fairly strong bias towards bigrams you haven't seen yet. (Of course, MCC's are a special case, since they're presumably patterns you want to practice).
This seems true to some extent. "qz" is an extreme case in that it's not even in the bigram frequency list I'm using, so only has (IIRC) some minimal chance of showing up at all. Even if q and z get pretty high weights, "qu" is much more likely to show up than "qz" because the bigram is so much more common. But again, I see this as a feature; if "qz" is slow, I'm fine with that; it's not something I type very often (except in this post )